air crash report
TRANSCRIPT
-
8/13/2019 Air Crash Report
1/133
AA2013-4
AIRCRAFT ACCIDENT
INVESTIGATION REPORT
FEDERAL EXPRESS CORPORATION
N 5 2 6 F E
-
8/13/2019 Air Crash Report
2/133
The objective of the investigation conducted by the Japan Transport Safety Board in
accordance with the Act for Establishment of the Japan Transport Safety Board and with
Annex 13 to the Convention on International Civil Aviation is to determine the causes of
an accident and damage incidental to such an accident, thereby preventing future
accidents and reducing damage. It is not the purpose of the investigation to apportion
blame or liability.
Norihiro GotoChairman,
Japan Transport Safety Board
-
8/13/2019 Air Crash Report
3/133
AIRCRAFT ACCIDENT INVESTIGATION REPORT
CRASH DURING LANDINGFEDERAL EXPRESS CORPORATIONMCDONNELL DOUGLAS MD-11F, N526FENARITA INTERNATIONAL AIRPORTMARCH 23, 2009
-
8/13/2019 Air Crash Report
4/133
SYNOPSISSummary of the Accident
On March 23 (Monday), 2009, about 06:49local time*1, a McDonnell Douglas MD-11F,
registered N526FE, operated by Federal Express Corporation as the scheduled cargo flight
FDX80, bounced repeatedly during landing on Runway 34L at Narita International
Airport. During the course of bouncing, its left wing was broken and separated from the
fuselage attaching point and the airplane caught fire. The airplane rolled over to the left
being engulfed in flames, swerved off the runway to the left and came to rest inverted in a
grass area.The Pilot in Command (PIC) and the First Officer (FO) were on board the airplane,
and both of them suffered fatal injuries.
The airplane was destroyed and the post-crash fire consumed most parts.
Probable CausesIn this accident, when the airplane landed on Runway 34L at Narita International
Airport, it fell into porpoising. It is highly probable that the left wing fractured as the load
transferred from the left MLG to the left wing structure on the third touchdown surpassed
the design limit (ultimate load).
It is highly probable that a fire broke out as the fuel spillage from the left wing caught
fire, and the airplane swerved left off the runway rolling to the left and came to rest
inverted on the grass area.
Th di t hi h th i l f ll i t th i h f ll :
-
8/13/2019 Air Crash Report
5/133
interpretation of the requirement at the time of type certification for the MD-11 series
airplanes.
Safety RecommendationsOn March 23 (Monday), 2009, about 06:49 JST (Japan Standard Time), a McDonnell
Douglas MD-11F, registered N526FE, operated by Federal Express Corporation as the
scheduled cargo flight FDX80, bounced repeatedly during landing on Runway 34L at
Narita International Airport. During the course of bouncing, its left wing was broken and
the airplane caught fire. The airplane rolled over to the left being engulfed in flames,
swerved off the runway to the left and came to rest inverted in a grass area on the westside of the runway.
The airplane approached with a high sink rate, with its autothrottle onamid strong
gusty winds and with unstable airspeed and attitudes. The late flare caused hard landing
and the airplane bounced. Large nose-down elevator input just before and during the
touchdown caused the second touchdown on the NLG with negative pitch attitude
developing into porpoising. Upon the third touchdown, the left wing structure fractured
because it surrendered to an overload transferred from the left MLG.
As a result of the investigation of this accident, the JTSB makes the following
recommendations to the Federal Aviation Administration of the United States of America
to take the following measures to prevent the recurrence of similar accidents.
Actions to Be Taken by the Federal Aviation AdministrationAl h h h MD 11 i l ifi d h i 14 CFR 25 721( )
-
8/13/2019 Air Crash Report
6/133
however, its not a fundamental solution. As the occurrences of vertical overload have been
reported after this accident, the measures taken so far are not considered to be satisfactory.
The JTSB recommends that the Federal Aviation Administration require the Boeing
Company to study the possibility of design change for the MLG support structure and
matters mentioned below in order to prevent the recurrence of similar accidents and
minimize damage to be caused by such accidents.
a. In order to reduce the occurrence of MD-11 series airplanes severe hard landing
and bounce in which an overload is transferred to the MLGs and their supporting
structure, the Boeing Company should improve the controllability and maneuver
characteristics by improving the LSAS functions, reducing the AGS deploymentdelay time and other possible means.
Possible improvement on LSAS functions may include: a function to limit large
nose-down elevator input during touchdown phase, which is a common
phenomenon in severe hard landing cases accompanied by structural destruction
for MD-11; and a function to assist bounce recovery and go-around in case of
bounce.
b. In order to help pilots to conduct recovery operation from large bounces and judge
the necessity of go-around, studies should be made to install a visual display and
an aural warning system which show gear touchdown status on MD-11 series
airplanes.
-
8/13/2019 Air Crash Report
7/133
Abbreviations used in this report are as follows:AC : Advisory Circular
AFS : Auto Flight System
AGS : Auto Ground Spoilers
ALPA : Airline Pilots Association
AMC : Acceptable Means of Compliance
AMM : Aircraft maintenance Manual
ATS : Auto Throttle System
CAS : Calibrated Airspeed
CAWS : Central Aural Warning System
CCP : Control Column Position
CFM :Company Flight Manual
CFR : Code of Federal Regulations
CST : Central Standard Time
CVR : Cockpit Voice Recorder
CWP : Control Wheel PositionDFDR : Digital Flight Data Recorder
EASA : European Aviation Safety Agency
ELF : Elevator Load Feel
EPR : Engine Pressure Ratio
FAA :Federal Aviation Administration
FCC : Flight Control Computer
-
8/13/2019 Air Crash Report
8/133
PF : Pilot Flying
PFD : Primary Flight Display
PIC : Pilot in Command
PIREP : Pilot Report
PM : Pilot Monitoring
PMI : Principal Maintenance Inspector
PNL : Positive Nose Lowering
POI : Principal Operations Inspector
PRD : Pitch Rate Damper
RA : Radio Altitude
RAAS : Runway Awareness Advisory System
RNAV : Area Navigation
RSPA : Research and Special Programs Administration
S/O
SPECI
: Second Office
: Aerodrome special meteorological reports
TRA : Thrust Resolver AngleUTC : Universal Time Coordinated
VASI : Visual Approach Slope Indicator
-
8/13/2019 Air Crash Report
9/133
TABLE OF CONTENTS
1. PROCESS AND PROGRESS OF THE INVESTIGATION ....................................... 1
1.1 Summary of the Accident ................................................................................................................. 1
1.2 Outline of the Accident Investigation ............................................................................................. 1
1.2.1 Investigation Organization .................................................................................................... 1
1.2.2 Representatives from Relevant States .................................................................................. 1
1.2.3 Implementation of the Investigation ..................................................................................... 1
1.2.4 Interim Report ......................................................................................................................... 2
1.2.5 Comments from Parties Relevant to the Cause of the Accident .......................................... 21.2.6 Comments from the Relevant States ..................................................................................... 2
2. FACTUAL INFORMATION........................................................................................ 3
2.1 History of the Flight ......................................................................................................................... 3
2.1.1 History of Flight According to DFDR CVR Records, Camera Images, Broadcast Images
and ATC Communications Records ....................................................................................... 3
2.1.2 Statements of Air Traffic Controllers................................................................................... 10
2.2 Injuries to Persons ......................................................................................................................... 10
2.3 Damage to the Airplane ................................................................................................................. 10
2.3.1 Extent of Damage ................................................................................................................. 10
2.3.2 Damage to Aircraft Components .......................................................................................... 10
2.4 Other Property Damage ................................................................................................................. 11
2.5 Personnel Information ................................................................................................................... 11
-
8/13/2019 Air Crash Report
10/133
2.8 Air Navigation Facilities ................................................................................................................ 20
2.9 DFDR and CVR .............................................................................................................................. 20
2.10 Information about Aircraft Wreckage and Traces ........................................................................ 21
2.10.1 Distribution of Wreckage and Traces .................................................................................. 21
2.10.2 Details of Damage ................................................................................................................. 23
2.11 Detailed Examination of Landing Gears ...................................................................................... 33
2.11.1 Examination of Left MLG .................................................................................................... 33
2.11.2 Teardown Examination of NLG and Left MLG Struts ...................................................... 33
2.12 Firefighting and Rescue Activities ................................................................................................ 37
2.13 The Airplanes Landing and Related Control Operations........................................................... 392.13.1 Procedure Included in the Chapter 7-1-5 Landing, the Companys Flight Manual........ 39
2.13.2 Summary of Pilots Statements about Controllability at Landing for the Same Type of
Aircraft ................................................................................................................................... 39
2.13.3 Landing Operations .............................................................................................................. 40
2.13.4 Bounce Recovery ................................................................................................................... 40
2.14 Tests and Researches ..................................................................................................................... 41
2.14.1 Windshear Warning .............................................................................................................. 42
2.14.2 Comparison of the Accident Landing with Previous 60 Landings .................................... 42
2.14.3 CAWS Call-out Interval ........................................................................................................ 44
2.14.4 Factors Linked to Pitch Attitude Variation......................................................................... 44
2.14.5 Longitudinal Stability Augmentation System (LSAS) ....................................................... 46
2.14.6 CG and Cockpit Height Change Below 50 ft RA ................................................................ 49
-
8/13/2019 Air Crash Report
11/133
3.1.1 Crew Qualification ................................................................................................................ 74
3.1.2 Physical Condition and Training ......................................................................................... 74
3.2 Airworthiness Certificate of the Airplane .................................................................................... 75
3.3 Relations to Meteorological Phenomena ....................................................................................... 75
3.4 Accident Scenario ........................................................................................................................... 76
3.4.1 From the First Communication Established with Narita Tower ...................................... 76
3.4.2 From 1,000 ft RA to Autopilot Off ........................................................................................ 76
3.4.3 From AUTOPILOT OFF Onward ........................................................................................ 77
3.5 The Load on Landing Gear at Touchdown and Structural Failure Sequence ........................... 83
3.5.1 Before Touchdown ................................................................................................................. 833.5.2 Around First Touchdown ...................................................................................................... 84
3.5.3 Around Second Touchdown .................................................................................................. 84
3.5.4 Around Third Touchdown ..................................................................................................... 84
3.5.5 Fracture of Left MLG Attachment Structure ..................................................................... 85
3.5.6 After the Fracture of the Left Wing ..................................................................................... 86
3.6 Fire .................................................................................................................................................. 86
3.7 Spillage of Fuel Due to Structural Destruction ........................................................................... 87
3.8 Fire Fighting and Rescue Operations ........................................................................................... 88
3.9 Active Use of Camera Images ........................................................................................................ 89
4. CONCLUSIONS ........................................................................................................ 90
4.1 Findings .......................................................................................................................................... 90
4.2 Probable Causes ............................................................................................................................. 92
-
8/13/2019 Air Crash Report
12/133
Attachment 1 CVR Records
Attachment 2 ATC Communications Records
Attachment 3 Camera Images from about 50 ft RA to Stoppage of Airplane
Attachment 4 Fire Fighting
Attachment 5 Working Conditions of the PIC and the FO
-
8/13/2019 Air Crash Report
13/133
1. PROCESS AND PROGRESS OF THE INVESTIGATION1.1 Summary of the Accident
On March 23 (Monday), 2009, about 06:49local time, a McDonnell Douglas MD-11F,
registered N526FE, operated by Federal Express Corporation as the scheduled cargo flight
FDX80, bounced repeatedly during landing on Runway 34L at Narita International
Airport. During the course of bouncing, its left wing was broken and separated from the
fuselage attaching point and the airplane caught fire. The airplane rolled over to the left
being engulfed in flames, swerved off the runway to the left and came to rest inverted in a
grass area.The Pilot in Command (PIC) and the First Officer (FO) were on board the airplane,
and both of them suffered fatal injuries.
The airplane was destroyed and the post-crash fire consumed most parts.
1.2 Outline of the Accident Investigation1.2.1 Investigation Organizationa. On March 23, 2009, the Japan Transport Safety Board (JTSB) designated an
Investigator-in-charge (IIC) and five investigators to investigate this
accident.
b. The JTSB also designated two expert advisers to look into the following
technical matters:
(1) For investigation into aircraft maneuver
-
8/13/2019 Air Crash Report
14/133
March 22 to 26, 2010 Teardown investigation of the nose landing gear (NLG) and the
left main landing gear (MLG)
October 26, 2009 to Flight maneuver analysis
April 30, 2011
October 26, 2009 to Structural analysis
November 2, 2011
November 8 to 9, 2011 Progress meeting
September 17 to 18, 2012 Exchange of opinions with the relevant States on the touchdown
operation of the MD-11 airplane
1.2.4 Interim ReportOn April 16, 2010, an interim report based on the fact-finding investigation up to that
date was submitted to the Minister ofLand, Infrastructure, Transport and Tourism (MLIT),
and made public.
1.2.5 Comments from Parties Relevant to the Cause of the AccidentComments were invited from the parties relevant to the cause of the accident.
As the PIC and the FO had been deceased, their comments were not available.
1.2.6 Comments from the Relevant StatesComments were invited from the relevant State.
-
8/13/2019 Air Crash Report
15/133
2. FACTUAL INFORMATION2.1 History of the Flight
On March 23 (Monday), 2009, a McDonnell Douglas MD-11F, registered N526FE,
operated by Federal Express Corporation (hereinafter referred to as "the Company"), took
off from GuangzhouBaiyun International Airport (hereinafter referred to as "Guangzhou
Airport") as the scheduled Flight FDX80 (a cargo flight) to Narita International Airport
(hereinafter referred to as "Narita Airport").
The flight plan of the airplane was summarized as below:
Flight rules: Instrument Flight Rules (IFR)Departure aerodrome: Guangzhou Airport
Estimated off-block time: 03:15
Cruising speed: 493 kt (913 km/h)
Cruising altitude: S1010*2
Route: Omitted A593 (airway) LAMEN (reporting
point)
Cruising speed 498 kt
Cruising altitude FL290
Route A593 (airway) FU (Fukue NDB) Y23
(RNAV route) PERRY (reporting point) Y231
(RNAV route) VENUS (reporting point)
Destination aerodrome: Narita Airport
-
8/13/2019 Air Crash Report
16/133
The airplane was flying with autopilot and auto throttle engaged.
06:41:39 Narita Tower informed the airplane that the runway in use was 34L
and its landing sequence was second. It also informed the airplane
that the wind was 320 at 28 kt (the maximum 40 kt and the
minimum 20 kt) and instructed the airplane to continue its
approach.
06:41:40 The FO instructed the PIC to perform a before-landing checklist.
06:41:51 The airplane read back the approach instructions to Narita Tower.
06:42:14 The FO asked the PIC about a speed additive to Vref (reference
landing speed) to get Vapp (approach speed). The PIC gave a value of10 kt as the additive. Vapp was set at 164 kt (Vref plus 10 kt).
06:42:32 The auto ground spoiler (AGS) was armed.
06:42:40 The flight crews confirmed AUTO BRAKE at MEDIUM and
GEARDOWN FOUR GREEN.
06:43:57 Narita Tower provided the pilot report (PIREP) to the airplane that
there was a windshear of plus or minus 15 kt on the final approach
toward Runway 34L below 2,000 ft and added that the surface wind
was 320 at 23 kt with the maximum 34 kt and the minimum 15 kt.
06:44:11 Responding to the wind information, the crews exchanged a
conversation to the effect that the situation was not so bad due to
near head wind.
06:44:50 The PIC advised the FO to check the windshear guidance and the FO
-
8/13/2019 Air Crash Report
17/133
within plus or minus 5 from 330
The vertical acceleration was 1.4 G at about 800 ft and 700 ft, and 0.4 G at
about 450 ft. But these vertical acceleration was not abrupt in nature.
The longitudinal acceleration was between +0.2 G and -0.1G, while its lateral
acceleration was 0.1 G.
06:46:53 Central aural warning system (CAWS) called 1,000 ft.
06:47:10 The PIC said, Yee haw rideem cowboy.
06:47:17 The runway awareness advisory system (RAAS) called an approach
to Runway 34L.
06:47:40 CAWS called 500 ft.It was flying with CAS 179 kt, heading 331, pitch angle 1.8. Its
Glide Slope (GS) deviation was 0.09 dot and Localizer (LOC)
deviation was -0.03 dot from the instrument landing system (ILS)
course.
06:47:42 The PIC reconfirmed the runway 34L and called stabilized
(followed by the recording of the PIC s laughter and the FOs
response.)
It was flying with CAS 166.5 kt, heading 331. The pitch angle was
decreasing from 0.7 to -1.1. ILS deviation was GS 0.16 dot and LOC
-0.03 dot.
c. From AUTO PILOT OFF to 50 ft RA (Neutral column position was -2 during
landing phase.)
-
8/13/2019 Air Crash Report
18/133
At 06:48:15, CAWS called out
50 ft with RA 48 ft, CAS 161 kt,
heading 333, pitch angle 1.1, roll
angle -1.4 to -1.1, and ILS
deviation GS -1.11 dot, LOC -0.04
dot. From just before this moment
until touchdown, the EPR was
decreasing from 1.3 and reached
1.0 at the time when the altitude
was less than 20 ft.At 06:48:17, the altitude
dropped below 30 ft. The control
column was slightly pushed forward
(and the corresponding CCP
changed from -2.5 to -0.7, and to
-3.0), but this column input was
not a distinctive one. The control
wheel was also steered to the left.
When the CCP was -3.0, the CWP
was -14.4. At 06:48:18, when the altitude dropped below 20 ft, the pitch angle,
which had been stable at 1.1 until then, decreased temporarily to 0.7.
At 06:48:18, the control column was pulled (CCP from 0.8 to 1.5, to 1.2 and
(Note) The figure at the upper left corner of theindicates the number given to the 61 pieces ofpictures From about 50 ft Radio Altitude to Stoppage
of aircraftas shown in Attachment 3
1 Passing through50ftRA6:48:15.6
6 Passing through30ftRA6:48:17.0
-
8/13/2019 Air Crash Report
19/133
The CAWS called out at every
10 ft from 50 ft to 10 ft RA. The
interval between 20 ft and 10 ft
was one second, while that of
other calls were less than one
second.
e. From the first touchdown to the
second touchdown
After the first touchdown, the
airplane bounced. The controlcolumn was pushed forward (CCP
from -4.9 to -6.3, and to -6.7 and
-5.7) for a very short period just
before and after the touchdown,
while the pitch angle continued to
decrease from 4.6 to -1.8 during
the corresponding period. When
the airplane reached the highest
point during the bounce, the pitch
angle was 1.1. According to a
Camera Image, at around
06:48:21 when the MLGs left the
21 AGS deployed 6:48:21.0
22 1stbounce 6:48:21.2
24 2ndtouchdown on NLG 6:48:21.8
-
8/13/2019 Air Crash Report
20/133
on the MLGs, the ground spoilers
were retracted, and the pitch
angle increased to 6.7 in about
one second (from 2.5 to 3.5 to
4.6 to 5.6 to 6.3 and to 6.7).
The control column forward input,
which temporarily decreased (to
-2.6), increased again (from -2.6
to -4.3 to -7.5).
During the second bounce,from 06:48:24 to 06:48:25, the
control column forward input
continued (-7.7 to -7.4 to -7.6),
and the pitch angle gradually
decreased (6.7 to 6.3 to 5.6 to
4.9 to 3.5 to 2.5 to 0.7).
At 06:48:25, the EPRs of the
No. 2 and No. 3 engines increased
to 1.1 from 1.0. When the pitch
angle decreased to 2.5, the
airplane reached the highest
bouncing point (16 ft). From
37 Highest bouncing point 6:48:25.2
39 Start of nose down 6:48:25.8
31 Pitch angle: at about6.7 6:48:23.6
-
8/13/2019 Air Crash Report
21/133
increased again to 2.98 G. The
maximum longitudinal
acceleration which was recorded
was 0.39 G backward, while the
maximum lateral acceleration
was 0.5 G to the left.
According to a Camera Image,
when the NLG and the Left MLG
touched down and the airplane
sank, the left main wing greatlybent downward. When the Right
MLG touched down, the fuselage
was steered to the right. The left
wing had been located to unusual
position to the fuselage.
After the MLGs touched
down, the pitch angle increased
again (from 3.9 to 6.0 to 8.1 to
10.2 to 11.2), while the roll
angle to the left also increased
(from -3.5 to -4.2 to -12.7 to
-26.7). The pitch angle decreased
48 Roll to the left 6:48:28.2
59 The airplane completelyinverted
6:48:31.1
51 Fire break out 6:48:29.0
-
8/13/2019 Air Crash Report
22/133
2.1.2 Statements of Air Traffic Controllersa. Narita Tower Air Traffic Controller
The controller assumed his duties at the station at about 06:30. He started
his jobs with communications with flight JAL710, two flights before the flight
FDX80, and he provided the windshear information then obtained to the following
flight NCA037. Because FDX80 had already switched to his frequency, the
controller thought that its flight crew had heard the information, but to be on the
safe side, he broadcast the information. When FDX38, which was the follow-on
flight after FDX80, called the controller, he provided the same windshear
information to FDX38. After the NCA037 landed, he asked its flight crewaboutthe windshear conditions and he read back the information so that it may be
heard by the crew of FDX80. Later, he issued a landing clearance to FDX80 with
new QNH.
There was nothing unusual in communications with the airplane, but because
he felt the airplane was in a level flight near the D taxiway before touchdown, he
thought it may go around, but it landed. The situation after that was the same as
shown in broadcasted TV images.No windshear warning had been issued.
b. Deputy Chief Air Traffic Controller
FDX80 was approaching on the final approach toward Runway 34L and it
didnt appear unstable, compared to the airplanes approaching on the final toward
Runway 34R. He tried to actively obtain the PIREP data and provide them to the
-
8/13/2019 Air Crash Report
23/133
2.4 Other Property Damagea. Seven runway lights (including Secondary Cable broken at four points) were
broken.
b. Two runway centerline lights were broken.
c. Three centerline lights on the highspeed taxiway were broken.
d. There were traces of scratches and scoops on the runway surface and the landing
area.
e. Spilled fuel and oil as well as the fire damaged the runway surface and the landing
area.
f. The holding pond and the soil above buried pipeline etc were contaminated withspilled fuel and oil.
2.5 Personnel Information2.5.1 Certificates and Flight Hoursa. PIC (Male, Age 54)
Airline transport pilot certificate (airplane) April 4, 1989
Type rating for McDonnell Douglas MD-11 June 4, 2000Class 1 aviation medical certificate
Validity September 30, 2009
Total flight time 8,132 hr 00 min
Flight time in the last 30 days 52 hr 26 min
Total flight time on the type of aircraft 3,648 hr 11 min
-
8/13/2019 Air Crash Report
24/133
08:28 local time (March 14, 16:28 UTC).
2.5.2.2 Flights from Anchorage to the Occurrence of the Accident and Previous72-Hour Activities for the PIC and FO
The PIC and the FO were paired for sequence of flights from Anchorage consequently
ended up in the accident.
a. The show-up time at Anchorage International Airport was at 09:55, March 15 local
time (March 15, 17:55 UTC).
b. They departed Anchorage at 10:48 local time (18:48 UTC) the same day and
arrived at Narita Airport at 11:16 local time (02:16 UTC), March 16.The flight duty hour was 8 hr 51 min and the flight time was 7 hr 28 min. The
rest layover time was 34 hr 44 min.
c. On March 17, they departed Narita Airport at 21:59 local time (12:59 UTC) and
arrived at Guangzhou Airport at 01:45 local time, March 18 (17:45 UTC, March
17).
The flight duty hour was 6 hr 15 min and the flight time was 4 hr 46 min. The
rest layover time was 26 hr 00 min.
d. On March 19, they departed Guangzhou Airport at 03:58 local time (March 18,
19:58 UTC) and arrived at Penang International Airport via Kuala Lumpur at
09:41 local time (01:41 UTC).
The flight duty hour was 7 hr 26 min and the flight time was 4 hr 40 min. The
rest layover time was 28 hr 19 min.
-
8/13/2019 Air Crash Report
25/133
UTC). According to a family member of the PIC who received his telephone call
from Guangzhou Airport, he sounded normal.
(See Attachment 5-1 and 5-2: Working Conditions of the PIC and the FO)
2.5.3 Medical Historya. PIC
He started to feel pain in the sacrum in August 2008 and took extended sick
leave for treatment from October 2008 to February 2, 2009. He submitted a
medical certificate on January 29, 2009 stating that he had made a full recovery.
He was on medication for cholesterol and hyperlipidemia and both drugs areauthorized in the aviation medical examination.
b. FO
He had an ankle bone fracture in 1995 and received a knee surgery in 2002.
He had passed the aviation medical examination.
He was on medication for hyperlipidemia and the drug is authorized in the
aviation medical examination.
2.5.4 Physical Conditions of PIC and FOThe physical conditions of the PIC and the FO during the flight from Clark
International Airport to Guangzhou Airport as the flight leg just before the occurrence of
the accident was as follows, according to the two captains who boarded the airplane on the
cockpit jumpseats:
-
8/13/2019 Air Crash Report
26/133
his physical condition was fine.
The FO usually slept for eight hours when he was off duty for an extended
period of time. He had no sleep disorder like insomnia, but when he returned
home from the trip to Asia, he apparently had difficulty in sleeping for one or twodays, perhaps because of changes in the sleep pattern during his job.
About 45 minutes before the occurrence of the accident, the FO told the PIC that they
knew they were both tired, and the PIC replied yes. Then the FO told the PIC that, if it
was real quiet, make some noise, and the PIC replied yes.
2.5.5 Possession of MedicinesThe following medications were found from the PICs personal effects, but no
medicines were found from those of the FO:
Zolpidem A non-benzodiazepine sleeping medicine (a sleep inducing
medicine with very short-term effect; effects carried over to the
next morning are limited)
Sonata A non-benzodiazepine sleeping medicineWatson 825 Acetaminophen/oxycodone (Pain-killer belongs to combined
medicine of medicine-class narcotic analgesic)
Mylan 345 A benzodiazepine anti-anxiety medicine (long-term type)
A benzodiazepine general anesthetic medicine (very short-term
type)
Excedrin Back and Body
-
8/13/2019 Air Crash Report
27/133
c. Criteria of Medicine Use
(1) The Federal Aviation Administration (FAA) allows a pilot to fly after
taking acetaminophen, aspirin and/or ibuprofen provided there are no
side-effects and the underlying medical condition doesn t interfere withthe safe performance of flying duties.
(2) The FAA doesnt allow a pilot to take oxycodone, diazepam (Valium), or
temazepam on a regular basis; however, the FAA allows a pilot to use
these medications on an occasional basis (defined as less than twice a
month).
This guidance presumes that there are no prolonged side-effects
from the medication and the underlying condition for which the
medication was taken does not interfere with the safe performance of
flying duties.
Once taken, a pilot has to wait at least five dosing intervals of the
medication before operating an aircraft.
2.5.7 Periodic Training and Othersa. PIC
He missed an opportunity of scheduled training during his extended sick
leave. But he received requalification training when his sick leave ended. In the
proficiency check on February 15 and 16, 2009, he performed landings, go-arounds
and other subjects in cross winds of 25 kt, and he successfully passed the check.
He received Bounce Recovery Training on August 20, 2006.
-
8/13/2019 Air Crash Report
28/133
2.6.2 EnginesType : Pratt and Whitney PW4462 turbofan engines
Engine position No.1 No.2 No.3
Serial number P723953 P723836 P723955
Date of manufacture December 1992 November 1990 December 1992
Total time 51,219 hr 53,828 hr 35,165 hr
Total cycles 7,046 8,073 5,696
The total time since the last periodical check (A check (every 250 flight hourson
March 21, 2009) was 7 hr 00 min.
2.6.3 History of Airplane Maintenance and Repair WorkThe airplane was one of the Companys McDonnell Douglas MD-11F fleet. The
airplanehad been maintained under a maintenance program authorized by the FAA. The
maintenance records show that the airplane underwent a C check (every 6,000 flight hours
or every 18 months) on November 9, 2007 with its total flight time of 40,262 hours. The
next C check was scheduled on May 7, 2009. The airplane received an A check (every 250
flight hours) on March 21, 2009 with its total flight time of 40,760 hours. The maintenance
records covering the last 30 days before the accident included a carried-over repair item. A
temporary repair work was done with aluminum alloy tape to cover a hole with a diameter
-
8/13/2019 Air Crash Report
29/133
2.6.6 Flammable Liquids on BoardThe airplane had about 400 kg of onboard flammable liquids. Major liquids are as
follows:
Polysilazane 75 units of 5.0-liter containerEthanol 2 units of 7.5-liter container
2.6.7 Natural Frequency of Fuselage StructureThe airplane manufacturer stated that the first bending frequency in the longitudinal
axis of the fuselage structure for the McDonnell Douglas MD-11 Series is approximately 3
Hz.
2.7 Meteorological Information2.7.1 General Weather OutlookThe Asia-Pacific surface analysis chart for 21:00 on March 22 depicted that a low
pressure system at 990 hPa in Hokkaido was moving to the northeast at 35 kt, while a cold
front from this low pressure system was extending to the middle of Honshu. Another low
pressure system at 996 hPa was moving to the east-northeast in the Tokai region with thecold front extending to the sea south of Japan. The areas around Japan were covered by a
pressure trough. Meanwhile, a high pressure system near Lake Baikal at 1,034 hPa was
moving to the southeast at 15 kt. Its influence spreadtoward the south, from the Yellow
Sea to the East China Sea.
The Asia-Pacific surface analysis chart for 09:00 on March 23 depicts that the low
-
8/13/2019 Air Crash Report
30/133
Altimeter setting (QNH) 29.49 inHg
06:30 Wind direction 320, Wind velocity 26 kt, Gust 40 kt / 13 kt,
Visibility 10 km or more,
Clouds Amount FEW, Type Cumulus, Cloud base 2,000 ft,Temperature 12 C, Dew point -2 C,
Altimeter setting (QNH) 29.52 inHg,
Windshear Runway 34L,
P/RR (pressure / rising rapidly)
06:50 Wind direction 310, Wind velocity 27 kt, Gust 39 kt / 16 kt,
Visibility 10 km or more,
Clouds Amount FEW, Type Cumulus, Cloud base 2,000 ft,
Temperature 12 C, Dew point -2 C,
Altimeter setting (QNH) 29.56 inHg,
P/RR (pressure / rising rapidly)
On the day of the accident, strong northwesterly winds were blowing. It died down late
at night.
2.7.3 Wind Observed by Doppler LIDAR (Light Detection and Ranging)A Doppler LIDAR is a system which observes the wind velocity, wind velocity
variation and other parameters by monitoring floating aerosol (dust particles in the
atmosphere, etc.). It rotates its head scanning laser beams at prescribed elevation angle
and bearing (Elevation angles of 1, 2, 3and 45in all directions, and elevation angles
-
8/13/2019 Air Crash Report
31/133
Doppler velocity observation data between 06:40:45 and 06:48:08 depicted
that a cold color area, representing a wind component toward the LIDAR site,
prevailed on the northwestern side of the LIDAR and a warm color area,
representing a wind component away from the site, prevailed on the southeasternside.
According to the analytical chart for 06:40:45, northwestern winds exceeding
15 m/s were observed at the center, and as time passed the winds at the center
weakened to 10-15 m/s. Around the final approach course of Runway 34L, 15-20
m/s of winds were observed, and the velocity remained unchanged.
No shear line was observed around the final approach course of Runway 34L.
No low level windshear information was issued, either.
b. Wind velocity variation and TURB
The wind velocity variation data for 06:40:45 and 06:48:08 depicted that the
number of red dots with the variation of 7 m/s or more, most of which prevailed
over the runway around the LIDAR site, became smaller as time passed.
Yellowish green dots with the variation of 4 to 5 m/s occupied most of the Runway
34L final. Yellow and orange dots with the variation of greater than 5 m/soccupied small portion of the area at 06:40:45, while red dots partially emerged at
06:45:40.
A red ellipse shows an area of TURB. Numbers of red ellipses were observed
over Runway 34L throughout the relevant time. On Runway 34L final, TURBs
(t07: maximum variation 9 kt) was detected near TH (threshold) at 06:45:40,
while it became smaller in size at 06:48:08.
-
8/13/2019 Air Crash Report
32/133
06:12 THA640 Fluctuating.
06:16 JAZ718 Below 10,000 ft plus minus 15 kt.
Rough air at 500 ft and below, till touchdown
06:19 JAL6524 Plus minus 10 kt below 1,000 ft.06:22 JAZ472 At 2,000, 32050 kt.
06:25 JAZ472 Below 1,000 ft plus minus 10 kt.
06:27 NCA007 Just a lot of rolling. Plus minus 15 kt.
06:32 NCA228 Plus 15, maybe sometimes 20, minus 10.
Below 1,000.
06:41 JAL710 Below 2,000 ft plus minus 15 kt.
06:46 NCA037 Really rough, plus minus 15 kt, below 1,000.
The predictive type and reactive type windshear warning devices were onboard flights
NCA007, NCA228 and NCA037; no warning was issued.
2.7.6 Relations between Information Mentioned in 2.7.5 and the Airplane in View ofATC Communications RecordsAt 06:42, the airplane called Narita Tower at 13 nm on the final to Runway 34L. The
airplane received the information that the wind then was 320 28 kt, with the maximum 40
kt and the minimum 20 kt.
Later, Narita Tower informed other airplane that the wind was 320 29 kt, with the
maximum 40 kt and the minimum 20 kt, and provided the PIREP from a Boeing 747
(Below 2,000 ft, wind velocity variations plus or minus 15 kt) at 06:41.
At 06:44 Narita Tower broadcasted the PIREP information from the Boeing 747 at
-
8/13/2019 Air Crash Report
33/133
CVR aboard the airplane, the device is described as CVR in this report.) made by GE
Aviation of the United States of America.
The DFDR retained data from the time when the airplane took off from Guangzhou
Airport to the accident at Narita Airport.The time was determined by correlating the VHF transmission keying signals in the
DFDR with the NTT speaking clock recorded on the ATC communications records.
The CVR was a 2-hour type and it retained voices at the time when this accident
occurred.
The past flight records left in the airplanesdata server were also used for this report.
DFDR
CVR
-
8/13/2019 Air Crash Report
34/133
(1) At 475 m from the threshold near the runway centerline, there were
traces of grease from the NLG.
(2) At 513 m to 596 m from the threshold near the runway centerline, there
were a hub cap, grommet, gasket and strap from the NLG.(3) At 565 m from the threshold near the runway centerline, there were four
pieces of NLG components (a band, wiring, etc.).
(4) At 596 m from the threshold on the left side of the runway, there was a
NLG wheel cap.
(5) At 700 m from the threshold on the left side of the runway, there was a
fragment of the left inboard flap.
c. At 795 m from the threshold on the right side near the runway centerline, there
were two nose tire rubber marks (9.6 m on the right and 8.85 m on the left)
followed by two tire wheel scratch marks (10 m on the right and 27.8 m on the
left).
d. At 814 m from the threshold near the runway centerline, there were scrape mark
(8.9 m by 0.9 m) by the left engine cowl.
e. At 875 m from the threshold on the right side of the runway centerline, therewere three pieces of NLG components (grommet, etc.).
f. At 877 m from the threshold near the runway centerline, there was a 79 m-long
scrape mark. Parallel with this mark on the right side of the runway centerline,
there was a 14.5 m-long scrape mark. On the left side of the runway centerline,
there were grooves and fragments of engine fan blades.
g. On A5 and A6 taxiways and near their crossings to the taxiways, there were
-
8/13/2019 Air Crash Report
35/133
2.10.2 Details of Damagea. Fuselage
The ceiling remained tied together from the nose to the tail section, but the
fire consumed most of the interior, from the cockpit to the tail. The fire also
consumed the lower outer skin of the fuselage around the main wings (the parts
facing upward due to the inversion) and most parts of the cabin including the floor,
leaving cargo fixture brackets scattered around.
The fire consumed the chemical fiber net and the synthetic rubber curtain,
Figure 1 Overview of the Airplanes Wreckage
-
8/13/2019 Air Crash Report
36/133
The fire consumed the wing root section, leading edge along with the inboard
slats near the wing attachment point to the fuselage, leaving leading edge and
slats. The trailing edge surface was sooted, but the fractured area of the rear spar
had retained its original shape. The double-plated rear spar, which consisted ofthe same thickness plate, was fractured almost at the same place. The fracture
formed a diagonal line linking the lower end near WS*3230.6 and the upper end
near WS197.2.Bends and cracks perpendicular to the fracture line were observed
at the mid section and the stringer at the base of the spar had span-wise cracks.
The outer skin in the area involved exhibited no vertical bending. (See Photo 3
and Figure 2)
-
8/13/2019 Air Crash Report
37/133
WS 197.2
WS 230.6
Cut surface after the accident (fuselage
Photo 3 Fractured Area of Left Wing (Fuselage Side)
Forward
Up
Right
-
8/13/2019 Air Crash Report
38/133
The No. 1 engine and the left MLG remained attached to the left wing
structure lay on its lower surface on the grass area to the east of the fuselage.
The shape of the broken face of the rear spar matched that of the fuselage
Photo 5 Fractured Left Wing
View from fuselage side
-
8/13/2019 Air Crash Report
39/133
-
8/13/2019 Air Crash Report
40/133
c. Right Wing
The right wing lay inverted attached to the fuselage. The fire consumed the
most part of the wing. The remaining parts from the fire were melted or sooted.
Photo 8 Right Wing
-
8/13/2019 Air Crash Report
41/133
lug, which is connected to the lower end of the landing gear, was torn apart. The
sooted left MLG strut was fully extended with no apparent damage. The surface of
the four wheels and tires were all burned and sooted. There were slight traces of
scratch on the two outer wheels. No. 5 and No. 6 tires were deflated without anyindications of burst. The torque link was undamaged. The MLG attachment
fitting had no damage, but in the lower part of fitting, there was dent caused by
the strut. (See Photos 9, 10 and 11, Figure 3)
(Source: AMM)Photo 10 Forward Trunnion Bolt (Fuse Pin)
-
8/13/2019 Air Crash Report
42/133
Photo 11 Left MLG
Crushed mark of the attachment fitting
-
8/13/2019 Air Crash Report
43/133
The center MLG remained attached to the inverted fuselage, but as the fire
consumed the fuselage interior, it somewhat drooped into the body. Borescope
image (See Photo 12) through the strut blow-out hole shows the smashed crush
tube*4, indicative of a strut bottoming. (See Figure 4)The sooted right MLG also remained and attached to the inverted fuselage.
The four tires were all sooted with traces of heat damage on the surface. The tires
remained inflated with no severe damages.
e. Engines
Photo 12 The Inside of the Blow-out Hole of Center MLG
-
8/13/2019 Air Crash Report
44/133
-
8/13/2019 Air Crash Report
45/133
2.11 Detailed Examination of Landing Gears2.11.1 Examination of Left MLGThe left MLG had no exterior damage except for the fire damage or that caused by the
GroundSpoiler/SpeedBrake Handle
Thrust Lever
Flap/SlatHandle
G/A Gate
Photo 16 Center Pedestal
-
8/13/2019 Air Crash Report
46/133
gas up to the regulated pressure. At landing, nitrogen gas is compressed from the extended
position and the fluid is forced to pass through the orifice that generates the resistance of
fluid flow so that the landing gear absorbs the landing force.
A teardown examination was carried out at the maintenance facility in the United
States of America for the NLG and the left MLG struts which had been installed in the
body sections which sank hard when the airplane made the third touchdown Later the
Figure 4 Mechanism of NLG/MLG Strut
(When extended)
Brow out Hole(Center MLG only)
Diaphragm(Center MLG only)
Cylinder
Nitrogen Gas
Ground Nut
Crush Tube(Center MLG
only)
Orifice andorifice support
tube
UpperBearing
Metering pin
Oil
(When compressed)
(Center MLG) (Other than Center MLG)Piston
-
8/13/2019 Air Crash Report
47/133
After the wheel tie bolts on both sides of the NLG were found to have been
loosened, the JAXA conducted a detailed examination on the wheels. It confirmed
that the tie bolts had become loose because the threads of the tie bolts and the
nuts had been damaged, then the nuts moved to loosen direction.The NLG wheel rims were crushed from one direction to be deformed into an
elliptic shape. It was estimated that a large load applied only once. Because the
shape of the deformation of the left and right rims was similar to each other, the
NLG tires on both sides had touched down almost at the same time. The condition
of the fracture surface of the rims showed that the fracture went from the bottom
to upward in Photo 19. This fact also showed that the NLG was exposed to a large
upward load (in the direction of being pushed up).
The findings are as follows:
(1) The rims were fractured with a upward load,(2) The wheel halves, which form the wheel hub, were extended outward and
deformed to opening direction, and
(3) The threads of the tie bolts and the nuts were fractured.
-
8/13/2019 Air Crash Report
48/133
A gap between outer and inner hubswhich form the left wheel
A part of inner hub deformed to the inside
Screw threads of centralart fractured in shear
The place where there isa gap between outer andinner hubs
Nuts at two placesdropped off
-
8/13/2019 Air Crash Report
49/133
b. Left MLG (part number: NRG6719-501 (D10-32-001-07, serial number:
BFGS00985)
Visual examination of the left MLG strut found no visible damage. Theexamination of the MLG components found no major damage with exception that
the lower lip of the upper bearing (the inner part that supports the upper end of
the piston) completely separated from the bearing halves at the place where it is
connected to the other components.
Fractured area
S f h i id
-
8/13/2019 Air Crash Report
50/133
07:18 The demolition fire engine started cutting open the rear part of the
cockpit from the western side of the airplane by the rescuers using
the engine cutter..
07:25 A Narita City firefighting rescue team started to enter the cockpit.07:30 The team secured an access route to the occupants, confirmed them
and continued to clear interior obstacles.
08:00 The PIC was rescued and transported to the hospital by ambulance.
08:16 FO was rescued and
transported to the
hospital by ambulance.
08:36 The fire was brought
under control. The efforts
were shifted to the ember
disposal and vigilance.
12:00 The ember disposal was
completed. The fire was
completely extinguished.
b. The fire-extinguishing activities
captured by Camera Images
At 06:49:28, the first fire engine
Picture 2 Fire Fighting Activities
1stfire engine Around 6:49:28
Opposite cameraStart of watering
Around 6:50:25
NoseFire
-
8/13/2019 Air Crash Report
51/133
2.13 The Airplanes Landing and Related Control Operations2.13.1 Procedure Included in the Chapter 7-1-5 Landing the Companys Flight ManualFollowing descriptions explain the landing and related control operations:
a. NORMAL LANDINGPlan to touch down 1500 ft from the runway threshold. The runway threshold
should disappear under the nose at about the same time CAWS announces "100."
Maintain a stabilized flight path through the 50 and 40 foot CAWS callouts
(unless sink rate is high). At 30' a smooth 2.5 degree flare should be initiated so as
to arrive below 10' in the landing attitude. Do not trim in the flare. Elevator back
pressure should be relaxed, and a constant pitch attitude should be maintained
from 10' radio altitude to touchdown.
The autothrottles switch to the retard mode at 50' RA. In the retard mode, the
throttles move to idle at a preprogrammed rate without regard to airspeed,
vertical velocity, or RA. The PF must maintain the appropriate glide path to
touchdown. If a deviation occurs from that glide path, the PF must override the
autothrottles to prevent retard.
At main wheel touchdown the autospoilers partially deploy, if throttle #2 is atidle. If throttle #2 is above idle at touchdown, AutoSpoilers and AutoBrakes may
not activate. Counter any pitch-up tendency associated with spoiler extension. Fly
the nose wheel smoothly to the runway. Avoid full elevator down input. If selected,
autobraking will begin shortly after spoiler deployment. When the nose wheel is
lowered to the runway, the spoilers will fully deploy.
(The rest is omitted)
-
8/13/2019 Air Crash Report
52/133
no engine in the tail, this type of aircraft is sensitive in the pitch correction and
therefore, steering must be made rather finely. No significant difference has been felt
in the control characteristics when the aircraft weight on landing is heavier and when
the weight is lighter. When strong winds are blowing, a sinking tendency may be feltfrom around when the ground effects emerge. When a gust is blowing, the speed
changes so greatly that the auto throttle cannot follow the situation. The pilots
concerned have not felt that this type of aircraft easily bounces.
Because these aircraft are equipped with LSAS (to be described later in 2.14.5),
the pilots have not felt any particular difficulties in controlling or any situation which
requires particular caution.
2.13.3 Landing OperationsWhen a line checker (hereinafter referred to as Captain A) of the Company was
interviewed about desirable landing operations with the cooperation of the investigation
organization (hereinafter referred to as NTSB) of the State of manufacture, comments as
summarized below were obtained.
The pilot would maintain the same attitude that had been used when followingthe glide slope and would not change pitch attitude when they heard the Central Aural
Warning System (CAWS) call out 100 feet. A pilot could begin the flare as early as 50
feet.
Light airplanes, heavy airplanes, and MD-10s all land differently. In a heavy
MD-11, it was best to begin the flare at 50 feet, but a light MD-10 was going to float if
a pilot flared that early The pilot had to think about how heavy the airplane was and
-
8/13/2019 Air Crash Report
53/133
Some pilots said they received tailstrike awareness and/or bounce recovery
training after their initial training and most indicated that they had at least
received some sort of tail-strike awareness or bounce recovery briefings since their
initial training. However, most pilots interviewed indicated that they onlyreceived tail strike awareness or bounce recovery simulator training during their
initial or transition training on the MD-11. As a result of the Narita accident, the
Operator is reevaluating the frequency of the tailstrike awareness and bounce
recovery training.
Captain A stated that in addition to providing the training in initial and
transition training, the Operator is immediately going to make tailstrike
awareness and bounce recovery simulator training either an annual or bi-annual
event.
b. Captain A
When asked to describe what she would do with the control column after the
initial bounce, the Captain said the procedure was to hold the nose at 7.5 degrees
and use the thrust as necessary to adjust the sink rate on landing or execute a go-
around if necessary. She did not know how high the accident airplane bounced butlooking at the video it looked like a high bounce, and she thought she would have
been bringing the throttle forward and executing a go-around at that time.
Bounce recovery raised a risk of a tailstrike, so she would not want to increase the
pitch too much.
Asked whether the flight crews would need to hold the control column a bit
aft to recover during a bounce, the Captain said yes, she would hold it a bit aft. If
-
8/13/2019 Air Crash Report
54/133
c. The flight characteristics of the airplane is reproduced based on the flight
characteristics data for the MD-11 series airplane provided by the airplane
manufacturer.
d. The flight environment (winds, atmospheric pressure, ambient temperature, etc.)at the time of the accident is reproduced based on the DFDR data and the flight
characteristics data for the MD-11 series airplane provided by the airplane
manufacturer. But vertical winds shall not be considered (the relevant value to be
set at zero).
e. The control inputs are produced based on the DFDR records. But the control input
data are corrected to simulate the airplane maneuver in the air (particularly its
attitude).
2.14.1 Windshear WarningAccording to the DFDR records, a
head wind component of the horizontal
winds which the airplane encountered
below 1,500 ft RA were about 60 kt at1,000 ft or above, about 50 kt from 1,000
ft to 500 ft, and about 25 kt near the
ground, meaning that there were the
wind velocity difference of about 25 kt
between the ground and 500 ft or more.
(Figure 6) The onboard windshear
-
8/13/2019 Air Crash Report
55/133
-
8/13/2019 Air Crash Report
56/133
d. The vertical acceleration just before and during touchdown was large, about 1.2 G
and 1.6 G, respectively.
e. The airspeed just before and during the touchdown is fast (166 to 169 kt CAS).
f. The time required for AGS deployment after touchdown was within the range ofvariations observed in the past landings. The AGS did not deploy to 60 the
maximum and they began to retract in 2 seconds after touchdown.
g. Other landings also had two positive vertical acceleration peaks upon landing,
just like those observed in this accident. These data suggest the airplane bounces
on those occasions. But in the accident landing, the height of bounce and the
change of vertical acceleration are larger than any other cases.
h. Forward control column movements just before and during touchdown were foundin some previous landings. However, these are different from this accident in that
the airplane attitude in these previous landings did not become negative pitch
angle after the first touchdown in spite of the forward control column input.
2.14.3 CAWS Call-out IntervalThe CAWS call-out is used as a timing reference to start flare or retard the thrust
lever.
The recorded call-out interval of CAWS in the CVR, with an interval of 10 ft from 50 ft
to 10 ft RA, and the average sink rate for every 10 ft are shown in Table 1. Suppose a pilot
starts flare at 30 ft, the follow-on sink rate decreases while the call-out interval becomes
longer. According to the FOM of the airplane manufacturer, the ideal sink rate should
finally be 2 to 4 fps (120 to 240 fpm) upon touchdown. In the accident, the call-out interval
-
8/13/2019 Air Crash Report
57/133
corresponds to the control column input.
c. Ground spoiler deployment generates nose-up pitching moment.
-
8/13/2019 Air Crash Report
58/133
2.14.5 Longitudinal Stability Augmentation System (LSAS)2.14.5.1 OutlineLSAS is a system developed and installed for MD-11 series airplanes to augment the
longitudinal stability. In December 1995, the pitch rate damper (PRD) function was added
to the system to increase the stability at high altitudes; moreover, a software package
(FCC-908) was introduced in May 2000 for the low altitude stability enhancement (LASE).
LASE mainly consists of the following three functions
a. The PRD function which mitigates changes in the pitch angle (applicable at high
altitudes and at low altitudes)
b. The pitch attitude protection (PAP) function which generates a nose down
command when the pitch angle approaches the threshold value (9.5) ontouchdown to provide tail strike protection.
c. The positive nose lowering (PNL) function which generates a nose down command
after the MLG touchdown and ground spoiler deployment to counter the pitch up
effect of ground spoiler deploymentThe NTSB concluded that FCC-908 software package will provide valuable
improvements in safety during MD-11 landings as stated in the Newark Accident*8 report.
The NTSB issued a recommendation (A-00-96) requiring the installation, within a year, of
the software package on all MD-11 airplanes. At present, the FCC-908 has been installed
in all the airplanes of the same series.
2.14.5.2 Influence of LSASLSAS has the PNL function which issues nose down commands following the MLG
-
8/13/2019 Air Crash Report
59/133
-
8/13/2019 Air Crash Report
60/133
-
8/13/2019 Air Crash Report
61/133
2.14.6 CG and Cockpit Height Change Below 50 ft RAThe cockpit of MD-11 series airplane is far ahead (by 31 m) of the CG. Hence, when the
pitch attitude variation is large, as was the case in this accident, it is possible that the
cockpit height changes differently from that of the CG. Hence, the height of the CG, cockpit
and MLG tires (height above runway) were examined with the flight simulation program
made by the airplane manufacturer. As the compression of the landing gear strut is not
taken into account, some MLG heights are shown below zero. (Figure 11)
The CG height and that of the MLG tire showed similar change; however, the changes
of the cockpit height showed differently as shown below.
a. The pitch angle started to increase about 1.5 seconds before the first touchdown
and this continued until the touchdown resulting in the smaller cockpit sink ratethan that of the CG. When the airplane touched down, the cockpit sink rate was
about 2 fps against the MLG tiresabout 7 fps.
b. At the time of the first bounce, the CG and MLG tires moved up about 4 ft.
Meanwhile, the cockpit sank about 9 ft not going up, due to the pitch angle
decrease.
c. About 3 seconds after the second touchdown, the cockpit reached the highest point
of about 40 ft. About 1 second later, the CG and the MLG tires reached the highest
point of about 33 ft and 16 ft, respectively.
-
8/13/2019 Air Crash Report
62/133
2.14.7 AGS DeploymentWhen the AGS is deployed, the lift decreases and would contribute to reduce the
height of a bounce. Therefore, the condition of the operation of the AGS was examined.
a. Required time for AGS deployment
(1) According to the airplane manufacturer, the estimated time from the
touchdown on the MLGs to the start of the deployment of the AGS is 1.25
seconds to be broke down as:
(a) Wheel spin-up detection ... 50 msec
(b) FCC relay activation . 200 msec
(c) Mechanically activation of the spoiler deployment
1,000 msec(2) The required time to start the AGS deployment (the time from the
increase in the vertical acceleration following the first touchdown to the
extension of the AGS by 10 or more) was examined for the landing in
this accident and in the previous 60 landings. (Figure 12)
The averaged time in the total 61 landings was about 1.4 seconds,
with a dispersion of about 2 seconds. The required time in the accident
landing was about 1.2 seconds, within a range of the observed dispersion.
Note that these values may have errors due to the data recording
intervals which are 0.5 seconds for AGS and 0.125 seconds for the
vertical acceleration.
The JTSB requested the Company via NTSB to provide the other
MD-11 airplanes data to survey the normal range of the required time for
-
8/13/2019 Air Crash Report
63/133
c. AGS retraction sequence
According to the airplane manufacturers manual, advancement of the No. 2
(central) engine thrust lever triggers the retraction of the deployed AGS. Because
the DFDR does not record thrust resolver angle (TRA), we used the calculated
Figure 12 Observed Required Time for AGS Deployment in the Previous
60 Landings
-
8/13/2019 Air Crash Report
64/133
-
8/13/2019 Air Crash Report
65/133
-
8/13/2019 Air Crash Report
66/133
-
8/13/2019 Air Crash Report
67/133
d. go-around executed during the second bounce; and
-
8/13/2019 Air Crash Report
68/133
g g
e. flight in a situation where the time required for AGS deployment is reduced, which
is considered to be effective for moderating a bounce after touchdown
We used the flight simulation program made by the airplane manufacturer.
2.14.10.1 Flare initiated at 30 ft RAThe airplane behavior was examined for the case where the airplane approaches in the
same manner as in this accident until 30 ft RA and starts flare at 30 ft RA. The
examination revealed the following: (Figure 15)
a. The DFDR records show that the sink rate did not decrease until around 10 ft RA
and it remained still 7 fps when the airplane touched down. Meanwhile, in thecase of flare start at 30 ft RA, the sink rate began to decrease around 20 ft RA and
the sink rate decreased to 2 fps at the time of touchdown.
b. When the flare started at 30 ft RA, the vertical acceleration on touchdown was
about 1.1 G. The bounce height after the touchdown was limited, and the vertical
acceleration on the second touchdown after the bounce was about 1.2 G. This
means if the flare had started in an appropriate manner at 30 ft RA, large bounces
after the touchdown, such as seen in the first and second touchdowns which
triggered this accident, would have not occurred.
-
8/13/2019 Air Crash Report
69/133
2.14.10.2 Nose-Down Column Input Held at Neutral Position Just Before First
-
8/13/2019 Air Crash Report
70/133
TouchdownAs described in 2.14.4, the elevators deflections corresponding to the control column
input, have the largest influence on the pitch attitude variations just before and after the
first touchdown and the nose-down motions during the two bounces. The airplane behavior
was examined in the case where the nose-down control column input just before the first
touchdown is limited to the neutral position (-2) and held it thereafter. The examination
revealed the following: (Figure 16)
a. The maximum pitch angle after the first touchdown was 5.0 and there was no
significant difference from that of 4.6 recorded in this accident.
b. Although the first bounce was larger than that of the accident, the second bouncewas quite small. No large pitch attitude variations occurred, either.
c. The airplane landed on the MLG first at the second touchdown with almost the
same sink rate of the second touchdown as that of the accident.
d. The maximum vertical acceleration (1.9 G) was observed upon the second
touchdown, but it was far less than the airplane tolerance limit which leads to a
destruction.
-
8/13/2019 Air Crash Report
71/133
2.14.10.3 Bounce Recovery Operation During the Second Bounce
-
8/13/2019 Air Crash Report
72/133
The airplane behavior was examined in the case where a bounce recovery is initiated
at 5 ft RA during the second bounce. The results are as follows: (Figure 17)
a. The pitch angle had already begun to decrease at the time of the recovery
initiation, but even at this stage the subsequent rapid nose-down motion could
have been avoided by applying large backward column input, and negative pitch
attitude could have been avoided.
b. Sink rate control by maintaining pitch angle 7.5 together with added thrust made it
possible to limit the vertical acceleration on the third touchdown to less than1.2 G.
However, the timing of the third touchdown is prolonged until 22 seconds
after the second touchdown by this bounce recovery. Considering that appropriateattitude and power control are necessary during that recovery operation and also
additional runway length is required, the bounce recovery may not be practical.
2.14.10.4 Go-around Operation During Second Bounce
-
8/13/2019 Air Crash Report
73/133
The airplane behavior was examined in the case where a go-around is initiated at 5 ft
RA during the second bounce. The results are as follows: (Figure 18)
a. The pitch angle had already begun to decrease at the time of the go-around
initiation, but even at this stage the subsequent rapid nose-down motion could
have been avoided with large backward column inputs, and negative pitch
attitude could have been avoided.
b. A go-around was possible before the third touchdown by increasing the thrust
simultaneously with a nose-up control column inputs mentioned above.
2.14.10.5 Reduced Required Time for AGS Deployment
-
8/13/2019 Air Crash Report
74/133
We examined the airplane behavior in the case where the required time for AGS
deployment is 0.61 secondsalmost the half of that observed in the accident. The results
are as follows: (Figure 19)
a. The AGS deployment during the first bounce generated nose-up moment and
restrained the nose-down motion. With reduced ASG deployment time, sooner
deployment is possible and this provides longer nose-up moment durations,
resulting in a large effect on nose-down restraint.
b. As the nose-down motion during the first bounce was restrained, the pitch angle
upon second touchdown was positive2.0, and the airplane touched down on the
MLGs. The loss of lift during the first bounce was also restrained since nose-downrestraint brought a higher angle of attack. This resulted in a smaller sink rate at
the touchdown6 fps, smaller than 9 fps recorded in this accident.
c. Because the second touchdown was made on the MLGs, the NLG did not receive a
bounce back from the runway surface resulting in very small nose-up moment
after the second touchdown. On the other hand, the sink rate was smaller than
that of the accident, so was the vertical acceleration on the second
touchdown1.5 G, smaller than that of the accident. These factors prevented any
bounce from happening after the second touchdown.
-
8/13/2019 Air Crash Report
75/133
2.15 Items Described in the Companys Manuals, etc.
-
8/13/2019 Air Crash Report
76/133
2.15.1 The Companys Flight Operations Manual (FOM)2.15.1.1 Chapter 2, General Policiesa. 2.5 FATIGUE
It is the crewmembers responsibility to be properly rested for each phase of
the trip. However, if circumstances prevent this, no FedEx crewmember should
feel pressured to fly when not properly rested. A crewmember who is fatigued
should immediately notify Crew Scheduling if unable to complete a trip.
b. 2.67 LEG SWAPPING
It has become customary for the Captain and First Officer to fly alternate legs.
This is an acceptable practice under normal circumstances. There are times,however, when the Captain may elect to conduct takeoffs and landings out of
sequence under unusual circumstances such as critical weather conditions or
complicated departure or arrival procedures. It is also important for the Captain
to consider landing currency when assigning flight legs.
(The rest is omitted)
2.15.1.2 Chapter 6, Arrivala. 6.45 STABILIZED APPROACH CRITERIA
(The forward part is omitted)
All flights must be stabilized by 1,000 feet above airport elevation when the
airport is IFR and by 500 feet above airport elevation when the airport is VFR.
The approach is considered stabilized when all of the following are met:
judgment this would create a greater hazard to flight safety.
b 6 47 GO AROUND PHILOSOPHY
-
8/13/2019 Air Crash Report
77/133
b. 6.47 GO-AROUND PHILOSOPHYThe decision to execute a go-around is both prudent and encouraged anytime
the outcome of an approach or landing becomes uncertain. FedEx considers the
use of the go-around under such conditions as an indication of good judgment and
cockpit discipline on the part of the flight crew.
2.15.2 The Company Flight Manual (CFM)CHAPTER 7-1-5 APPROACH
a. STABILIZED APPROACH 7-1-5-1Good landing are the result of good approaches. The majority of landing
accidents/incidents can be attributed to an unstable approach between the FAF
and touchdown. Control of the variables that result in a stabilized approach
should start prior to the FAF.
These variables are configuration, speed, rate of descent, power setting and
approach profile.
b. WIND ADDITIVE ON APPROACH 7-1-5-2Vapp is the greater of Vref+5 or Vref+wind additive. Wind additive is one half
of the steady state wind greater than 20 kt or full gust whichever is greater
(maximum 20 kt).
With ATS-OFF, the pilot manually maintains Vapp. With ATS-ON, the pilot
may apply wind additive by any method:
(Omitted)
The PAP sub-function will be active for both takeoff and landing.
New Positive Nose Lowering (PNL) Sub function This new sub function will
-
8/13/2019 Air Crash Report
78/133
New Positive Nose Lowering (PNL) Sub-function This new sub-function will
only be applied during landing phase, and will not affect AUTLAND operations or
Auto Flight System (AFS) autoland performance. The PNL sub function will apply
approximately 3 degrees of nose-down elevator command at main wheel spin-up,
at the same time that the FCC commands the Auto Ground Spoilers (AGS) to
extend. As the spoilers extend beyond 10 degrees, the second phase of PNL will
increase the nose-down elevator command to approximately 4 degrees. This
sub-function, in combination with the enhanced PRD, will control the elevator to
avoid aircraft nose-rise after touchdown, and to assist in de-rotation.
These LSAS changes will not affect the maximum override forces for LSAS, whichare dependent upon the position of the Elevator Load Feel (ELF) actuator for any
given airspeed. For the takeoff and landing phases, where PAP and PNL are
active, the control column forces required to override LSAS are approximately 10
to 15 pounds.
2.15.4 The Companys Flight Manual (CFM)Chapter 8-10A FLIGHT CONTROLS DESCRIPTION AND OPERATION:
a. SPOILER SYSTEMAuto Ground Spoilers (AGS)
After landing, all ten spoiler panels may be extended to maximum deflection
by automatic operation of the SPOILER handleIn order for this to occur, the
SPOILER handle must be armed and the flaps 31 degrees or more.
and a general dynamic analysis program ADAMS used in the past two events. The
analyses simulation was conducted using the following initial condition consistent with the
-
8/13/2019 Air Crash Report
79/133
analyses simulation was conducted using the following initial condition consistent with the
airplane at the 3rdtouchdown, and the time when the condition occurred was referred as
zero time.
a. The airplane weight, 405,120 lb with the CG at 31 % MAC.
b. The DFDR data (pitch angle, -4.6; left roll, 3.7; left yaw, 8.7; longitudinal
acceleration, 0.15 G backward; lateral acceleration, 0.13 G to the left; vertical
acceleration, 1.21 G upward (including a lift of 0.56 G).
The analysis model for this simulation has the following restrictions:
a. The data from the landing gear drop test is available up to 12 fps (the upper
design limit) sink rate. Estimated landing gear loads above 12 fps sink rateinclude errors in that there is no reference data available beyond the design limit,
where nonlinear effect comes out.
b. The behavior of the cylinder and piston of the strut after the bottoming is complex
so that the analysis model does not take this into account.
c. Structural failures are not taken into account.
The analysis showed that the left MLG strut bottomed at 0.2 seconds after the left
MLG contacted with the runway, considering the piston stroke. The right MLG strut
bottomed at 0.25 seconds. Both MLG was subjected to a load of at least 800,000 lb or more
and 950,000 lb maximum with errors included.
Based on the above load, it is estimated that the rear spar shear flow around the failed
area exceeded the fracture load in 0.2 seconds.
As this analysis assumes no structural failure the shear flow continued to increase to
The result of this analysis almost matched the structural analysis simulation
conducted by the airplane manufacturer
-
8/13/2019 Air Crash Report
80/133
conducted by the airplane manufacturer.
2.17 Landing Gear Design2.17.1 Landing Gear Design Criteria and Certification
The design criteria*9 applied to the MD-11 airplane requires the MLG system to be
designed so that, if it fails due to overloads during takeoff and landing assuming the
overloads to act in the upward and aft directions, the failure mode is not likely to cause the
spillage of enough fuel from any part of the fuel system to constitute a fire hazard. (14 CFR
25.721 (a))
One of the design method to satisfy with the requirement is to equip the MLG supportstructure with a fuse pin*10. In the design of the MD-11 airplane, the fuse pins had been
already installed in the MLG support structure.
However, the accident involved the MD-11 airplane in Newark in 1997 is concluded
that, as it is the case for this accident in which the opposite wing structure was destroyed,
a vertical (upward) overload transferred from the right MLG fractured the right wing
structure followed by fuel spillage. Regarding that the MLG was not separated by the
vertical overload, the Newark accident report states that MD-11 landing gear fuse pins
were designed and positioned to allow the landing gear to fail under loads in the aft
direction. The FAA explained, in its response to the NTSB safety recommendation
A-00-102, that in designing the landing gear to comply with 14 CFR 25.721(a), the MD-11
main landing gear was designed for an overload condition in which the drag load was the
primary component This landing gear was not designed to separate for a purely vertical
Compliance (AMC) for the European Certification Specification (CS) 25.721(a) which
correspond to 14 CFR 25.721(a). (The specification has also been revised, adding side loads
-
8/13/2019 Air Crash Report
81/133
correspond to 14 CFR 25.721(a). (The specification has also been revised, adding side loads
as an acting overload directions.)
The means of compliance with standards that are described in the guidance material
such as AMC and Advisory Circular of FAA illustrates the acceptable means. In addition,
other means are not excluded. The sole issuance of the advisory material will not lead to a
mandatory requirement to include the assumption of a primarily vertical overload, as
observed in the accident at Narita and Newark airports.
2.17.2 Criteria Applied to the Airplane (Excerpt)The following descriptions are included in14 CFR 25. 721:General
(a) The main landing gear system must be designed so that if it fails due to overloadsduring takeoff and landing (assuming the overloads to act in the upward and aft
direction), the failure mode is not likely to cause
(Omitted)
(2) For airplanes that have a passenger seating configuration, excluding pilots
seat, of 10 seats or more, the spillage of enough fuel from any part of the fuel
system to constitute a fire hazard.
2.17.3 Current Criteria and Interpretation Guidelines Revised by EASA (Excerpt)2.17.3.1 Descriptions in CS 25. 721 General (See AMC 25. 963 (d))(a) The landing gear system must be designed so that when it fails due to overloads
MD-11F, N611FE, operated by Federal Express, Inc., (FedEx) as flight 14, crashed
while landing on Runway 22R at Newark International Airport. The scheduled
-
8/13/2019 Air Crash Report
82/133
while landing on Runway 22R at Newark International Airport. The scheduled
cargo flight originated in Singapore on July 30 with intermediate stops in Penang,
Malaysia; Taipei, Taiwan; and Anchorage, Alaska. On board this airplane were
the captain, the first officer who had taken over the flight in Anchorage for the
final leg to Newark and three passengers. All five occupants received minor
injuries in the crash and during subsequent egress through a cockpit window. The
right wing separated from the fuselage just inboard of the wing MLG near WS264
followed by fuel spillage. The right engine and the right MLG were separated from
their attach points. A fire broke out.
The FDR data indicated that after the airplanes initial touchdown, it becameairborne and rolled to the right as it touched down again. The airplane continued
to roll as it slid down the runway, coming to rest inverted about 5,126 ft beyond
the runway threshold and about 580 ft to the right of the runway centerline.
b. Safety Recommendations
(1) A-00-092: to FAA: Convene a joint task force to develop a pilot training
tool to include information about factors that can contribute to structural
failures involving the landing gear, wings, fuselage, etc.(2) A-00-093: to FAA: Convene a joint task force to develop a pilot training
tool to provide a syllabus for simulator training on the execution of
stabilized approaches to the landing flare, the identification of
unstabilized landing flares, and recovery from these situations, etc.
(3) A-00-094: to FAA: Convene a joint task force to develop a pilot training
transport-category airplane designs based on the study results of
recommendation A-00-100.
-
8/13/2019 Air Crash Report
83/133
(11) A-00-102: to FAA: Conduct a study to determine if landing gear vertical
overload fusing offers a higher level of safety than when the gear is
overdesigned. If fusing offers a higher level of safety, revise 14 CFR Part
25 to require vertical overload fusing of landing gear. (The FAA had
earlier approved the designs which have no assumption of an overload
failure mode in which an upward load is outstanding, but in reply to this
safety recommendation, the FAA expressed its view that these designs
had satisfied airworthiness requirements under the interpretation at
that time, but they were not compatible under the interpretation as of2004.)
(12) A-00-103: to FAA: Require manufactures of CFR Part 23 and Part 25
airplanes and Part 121 operators to revise their hard landing inspection
and reporting criteria to account for all factors that can contribute to
structural damage; instruct principal maintenance and operations
inspectors assigned to Part 121 operators to ensure that these changes
have been made to operator maintenance manuals and Flight OperationsQuality Assurance exceedence monitoring programs.
(13) A-98-80: to RSPA (The Research and Special Programs Administration):
Require that air carriers transporting hazardous materials have the
means to quickly retrieve and provide consolidated specific information
of all hazardous materials on an airplane in a timely manner to
-
8/13/2019 Air Crash Report
84/133
Table 3 Hard Landing Cases with Structural Failures Involved for MD-11 Series Airplanes
(From 1993 to 2010)
Date Airport Operator Event
4/30/1993 Los Angeles Delta Air Line Bounced hard landing
8/19/1994 Chicago Alitalia Landing bounce and porpoise
7/31/1997 Newark FedEx Wing spar break and rollover
8/22/1999 Hong Kong China Airlines Wing spar break and rollover
5/22/2000 Taipei Eva Air Hard landing and go around
11/20/2001 Taipei Eva Air Bounce and nose landing gear (NLG) strike6/7/2005 Louisville UPS Hard NLG strike
3/23/2009 Narita FedEx Wing spar break and rollover
6/3/2009 Urumqi China Cargo Hard landing and tailstrike
9/6/2009 Khartoum Saudi Arabian Airlines Hard landing
9/13/2009 Mexico City Lufthansa Cargo Hard landing and NLG strike
10/20/2009 Montevideo Centurion Hard landing and main landing gear collapse
7/27/2010 Riyadh Lufthansa Cargo Hard landing and fuselage failure
9/22/2010 Kabul World Airways Hard NLG strike
(Source : NTSB)
b. Safety Recommendations to FAA:
(1) A-11-68: Require Boeing to revise its MD-11 FCOM to reemphasize high
accelerations at the beginning of the touchdown, and the airplane roll and pitch
attitude effects on the number of effective landing gears that absorb the landing
-
8/13/2019 Air Crash Report
85/133
energy. If the vertical acceleration at the beginning of the landing is less than 1G, the
potential energy that is normally accommodated by wing lift may instead be
transmitted into the landing gear.
D. Operators should evaluate the effects of pitch and roll and determine if a singular
landing gear may have absorbed significant loading. To determine if such loading has
occurred, use the landing Gs vs roll angle graph. If still undetermined, operator may
submit landing parameter to their Boeing Field Service Representatives of Boeing
Service Engineering for analysis.
3. ANALYSIS
-
8/13/2019 Air Crash Report
86/133
3.1 Airmen Competence Certificate and Others3.1.1 Crew QualificationThe PIC and the FO had held both valid airman competence certificates and valid
aviation medical certificates.
3.1.2 Physical Condition and Traininga. As described in 2.5.3 a, the PIC had pains in the sacrum, but he submitted a
medical certificate stating that he had made a full recovery.
b. As described in 2.5.7 a, it is highly probable that the PIC returned to his flightduties after having missed a chance of receiving periodic training during the
extended sick leave.
However, he received the requalification training and his skill was evaluated
as appropriate in the proficiency check.
c. As described in 2.5.7 b, it is highly probable that the FO received scheduled
training and passed the examination.
d. The description in 2.13.4 shows that the Company provided bounce recoverytraining to all MD-11 pilots after the Newark accident in 1997. It is highly
probable that this training was later provided only on the occasion of initial
training and transition training. In light of the personal history of the PIC and the
FO described in 2.5.7, it is highly probable that they had received bounce recovery
training in 2006 respectively
it could not be conclusively determined to what extent the fatigue influenced their